"Can I run a VM on Kubernetes?" is a solved problem — KubeVirt answers it. "Can I manage a fleet of them the way an operations team actually needs to?" is a bigger question, and it's what separates a proof-of-concept from something a platform team runs in production. Here's the full lifecycle that real VM management covers.
Golden images and flavors
Every new VM should start from an approved, patched baseline — a golden image — rather than an ad hoc disk someone configured by hand. Pairing images with standard flavors (predefined CPU/memory/disk sizes) keeps VM sprawl from turning into a spreadsheet of one-off configurations nobody can audit.
Snapshots and clones
A snapshot captures a VM's disk state before a risky change, so a bad upgrade is a rollback instead of an incident. Cloning turns a known-good VM into a template for standing up more of the same — a faster, more consistent path than re-provisioning from scratch each time.
Live migration
Nodes need maintenance, drain for upgrades, or occasionally fail. Migration moves a running VM off a node with minimal disruption, tracked as a first-class operation with progress visibility — not an SSH session and a prayer.
Backup and restore policies
Snapshots protect against a bad change; backups protect against losing the VM entirely. Backup and restore policies applied per VM — not bolted onto the hypervisor as an afterthought — mean recovery point and recovery time objectives are something the platform actually enforces.
SSH key management and console access
Operators need a way in when something goes wrong that isn't "share one root password in a chat message." Managed SSH keys and in-browser console access, both governed by the same RBAC as everything else, keep access auditable without slowing anyone down.
Access control and audit logging
Every one of the operations above — create, snapshot, migrate, restore — is a privileged action. Real VM management means every one of them is RBAC-checked and lands in an audit log, the same standard already expected of container workloads.
How AetherVirt covers the full lifecycle
AetherVirt implements every stage above as a tracked, permission-checked operation in one console: golden images and flavors, snapshot/clone/migrate with progress tracking, per-VM backup and restore policies, managed SSH keys, in-browser console access, and full audit logging — see the full platform capability list for the complete picture.